Salesforce Certified Data Architect Practice Test - Questions Answers, Page 18

The data architect should recommend building a batch job to move two-year-old records off platform, and delete records from Salesforce as an archiving solution. A batch job is a process that runs in the background and performs operations on large volumes of data in Salesforce. By building a batch job that moves two-year-old records off platform to an external storage system, such as Amazon S3 or Google Cloud Storage, and deletes them from Salesforce, the data architect can reduce the storage consumption and improve the performance of Salesforce org. Option A is incorrect because using a third-party backup solution to backup all data off platform will not free up any storage space in Salesforce, unless the data is also deleted from Salesforce after backup. Option B is incorrect because building a batch job to move all records off platform, and delete all records from Salesforce will result in losing all the current data in Salesforce, which may not be desirable or feasible. Option D is incorrect because building a batch job to move all restore off platform, and delete old records from Salesforce does not make sense, as restore implies restoring data back to Salesforce, not moving it off platform.

According to theSalesforce Connect Reportingblog post, one of the considerations for using Salesforce Connect for reporting is the OData callout limits per day. The blog post states that ''Salesforce Connect has a limit of 100,000 callouts per day. This limit is shared across all external data sources in your org. If you exceed this limit, you will receive an error message and no more callouts will be allowed until the next day.'' Therefore, a data architect should consider this limit before implementing Salesforce Connect for reporting.

According to theWhat is Master Data Management (MDM)?article, some of the data challenges that an MDM tool can solve are data duplication, data accuracy and quality, and data standardization. The article states that ''MDM solutions comprise a broad range of data cleansing, transformation, and integration practices. As data sources are added to the system, MDM initiates processes to identify, collect, transform, and repair data. Once the data meets the quality thresholds, schemas and taxonomies are created to help maintain a high-quality master reference.'' Therefore, an MDM tool can help NTO eliminate data duplication across different systems, improve data accuracy and quality by removing errors and inconsistencies, and standardize data formats and definitions for better integration and analysis.

According to theGet Started with Big Objectsunit on Trailhead, one of the use cases for custom big objects is to store and manage loyalty program data for customers. The unit states that ''From loyalty programs to transactions, order, and billing information, use a custom big object to keep track of every detail.'' Therefore, a custom big object can be used to capture the reward program data and display it on the contact record. Additionally, according to theBig Objects Implementation Guide, big objects can handle massive amounts of data (up to billions of records) and can be updated nightly from external systems using Bulk API or batch Apex. Therefore, a custom big object can meet the requirements of NTO's loyalty program scenario.

According to theUsing Mashupsarticle on Salesforce Developers, one of the techniques for deploying large data volumes is to use mashups to display historical records in Salesforce. The article states that ''Mashups are a way to display data from an external system within a Salesforce page without copying or synchronizing the data. Mashups use a combination of Visualforce, Apex callouts, and JavaScript code that runs in the browser. Mashups are useful when you want to display large amounts of read-only data that is stored outside of Salesforce.'' Therefore, a data architect should recommend this solution to meet the business requirement of UC.

According to theData Modelingunit on Trailhead, some of the key actions that should be done when data modeling in Salesforce are identifying data elements, mapping legacy data, and implementing legacy data model. The unit states that ''Before you start creating objects and fields in Salesforce, you need to identify the data elements that you want to store and work with. ... Next, you need to map your legacy data to Salesforce objects and fields. ... Finally, you need to implement your data model in Salesforce by creating custom objects and fields using declarative tools or Metadata API.'' Therefore, these are the correct actions for migrating a legacy client server app to Salesforce.

According to theGet Started with Custom Metadata Typesunit on Trailhead, one of the use cases for custom metadata types is to define custom charges for an accounting app. The unit states that ''Say that your org uses a standard accounting app. You can create a custom metadata type that defines custom charges, like duties and VAT rates. Then you can write some Apex code that calculates the total amount due for each invoice by using the metadata from your custom metadata type.'' Therefore, a similar approach can be used to maintain the pricing criteria for a custom pricing engine in Salesforce.

According to theHow to avoid row lock or race condition in Apexblog post, one of the ways to prevent race condition in Apex is to use the FOR UPDATE keyword in SOQL statements. The blog post states that ''We need to lock the records on which we are working such that other batches or threads will not be having any effect on them. How can we lock a record, then? We need to make use of FOR UPDATE keyword in the SOQL query.'' Therefore, a data architect should recommend this solution to guarantee that records are not being updated at the same time by different processes.

According to the Salesforce documentation1, ownership skew occurs when a large number of records (more than 10,000) are owned by a single user or queue. This can cause performance issues and lock contention when multiple users try to access or update those records. To prevent ownership skew, some of the recommended practices are:

Assign a default user as the owner of the records and do not assign any role to the default user (option C). This way, the records will not be visible to other users in the role hierarchy and will not cause sharing recalculations.

Keep users out of public groups that can be used as the source for sharing rules (option B). Sharing rules based on public groups can cause excessive sharing calculations and lock contention when many records are owned by a single user or queue.

Assign a default user as the owner of the records and assign the top most role in the hierarchy to the default user (option E). This way, the records will be visible to all users in the role hierarchy, but will not cause sharing recalculations or lock contention.

Assigning a default user as the owner of the records and assigning a role in the hierarchy (option A) is not a good practice, as it can cause sharing recalculations and lock contention when the role is updated or moved. Assigning ''view all'' permission on profile to give access to the records (option D) is also not a good practice, as it can bypass the security and sharing model and expose sensitive data to unauthorized users.